The gambler's fallacy and the hot hand in the same trader

[Shadow Trader]

After four losses in a row, a win feels due. After four wins in a row, the streak feels unstoppable. Both thoughts are mistakes, and they pull in opposite directions.

If trades are independent, the probability of the next win is unaffected by recent outcomes. A coin that landed on heads four times is still 50-50.

The gambler's fallacy leads to bigger bets after losses, expecting a reversal. The hot hand fallacy leads to bigger bets after wins, expecting continuation.

In trading, outcomes are not perfectly independent, since market regimes persist. A strategy that suits the current regime will win more for a while, and one that does not will lose more. That is a different story from luck owing something.

Fixed risk per trade, regardless of recent outcomes, sidesteps both fallacies. Reviewing performance by regime, not by streak, gives better information.

How has position sizing behaved after streaks in the records here?

[Amanhasnoname]

I recognised both of these in my own records once I started marking position size next to each trade. After a few losses I used to bump size slightly because I felt a winner was coming, and after a good run I did the same thing because I felt sharp. Two different stories, same outcome, more risk at exactly the wrong times.

When I looked at three months of trades, my biggest losing days were almost all ones where size had drifted up after a streak in either direction. My actual win rate after streaks was basically the same as overall, so the extra size was pure exposure with nothing behind it.

What fixed it for me was taking the decision away completely. I set my lot size once a week based on the account balance on Monday and it does not change until the following Monday, no matter what happens in between. I agree with your point about regimes though. I now tag each day as quiet or active and review results that way, which tells me far more than any streak ever did.